[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[McIDAS #ICA-762524]: mcidas-V

Hi Barry,

> Thanks for helping Randy,

No worries.

> and good to see you in Madison last week!

I regret that I did not have enough time to visit with everyone!  I did get
a chance to talk to Dave about the correct procedure for getting my bug
fixes into the McIDAS CVS repository and to Kevin B. about some changes
in XCD code that allow GRIB decoding/filing to work on big-endian systems,

> Regarding the list of changes for the 2007a Fastrack being incomplete, I
> assume you're referring to the list at
> http://www.ssec.wisc.edu/mcidas/software/x/changes2007a.txt. Is that
> correct?

Yes, that is the page I was referring to.

> I checked with Becky, Rick and Dave, and they were surprised that the
> list could be wrong.

The list is incomplete.  Please remember that I do not do an installation
of the full v2007a distribution.  Instead, I need to get a list of _all_
of the files that were changed so that I can incorporate the changes into
my distribution.  Given the large number of bugfixes/additions/modifications
that I make in the process of creating Unidata McIDAS-X/-XCD, it is not 
for me to use full Fastracks; I routinely use the Incremental upgrades to
get the list of files that have been updated.

> Becky also mentioned the same thing that I did last
> week: that perhaps the updating of the makefile on 12oct was related?

Yes.  'makefile' was and still is missing from the list of files that were 
for the upgrade.

> (I note from the weblogs that yakov.unidata.ucar.edu initially downloaded
> 2007a on 09oct.)

Yes, but I downloaded it again to my home development system this past weekend,
so I am working with whatever version was available on Saturday, October 20.

> We'll take a look and see if we can find any problems. If you do the
> "slogging through the full v2007a release" :-) and find anything, please
> let us know.

There was no problem with routines not being upgraded.  The problem was with
the listing of the routines that had to be changed.

Here is the list of routines modified for the SSEC v2007a Fastrack:

mclisten.c 1.16 
mclconst.h 1.2
sendimage.c 1.3
sendimage.h 1.3
imgremap.f 1.104
mtstaget2.f 1.13
mtstadir2.f 1.12
nvxgeos.dlm 1.3
constants.h 1.6
build-dmakefiles.pl 1.132
DMake.unix 1.49

My code diving shows that the following routines were needed:

makefile          <- not included in the list of modified routines
mclisten.h        <- not included in the list of modified routines

Also, the list of changes did not indicate that 'constants.h' had been renamed 
'mclconst.h' along with being modified.  The renaming of the routine was
one of the changes reflected in 'makefile'.

After figuring out the correct list of files that had been modified, I was able
to incorporate the changes in Unidata McIDAS to create my v2007g release.  My
release also contained numerous code changes to fix a problem where the wrong
type parameter was being passed as the third parameter to the McIDAS C routine
Mcdaytimetosec (in daytime.c).  I also discovered the likely reason that
there were so many routines using the incorrect type for that variable:  the
man page for Mcdaytimetosec does _NOT_ match the function declaration in
daytime.c.  This is a HUGE problem when one builds in 64-bit mode!  In fact, it
was the cause of strange behavior I reported on the serving of GVAR data
off of unidata2.ssec.wisc.edu.

Here is the full list of C routines that call Mcdaytimetosec with indications
of the type of parameter they were using as the third argument:

Routine         Pkg   Orig Declaration(s)                Unidata Revisions
daytime.c       X                                        changed help
servutil.h      X     time_t (FILELIST struct)           int
wwdisp.c        X     int *
wwlist.c        X     int *, static int *
wwmisc.c        X     int *
gvaradir.cp     X     time_t *                           int *
gvaraget.cp     X     time_t *                           int *
lv1butil.c      X     time_t *                           int *
mcts.c          X     int * (TS_TIME struct)
modsks.cp       X     static time_t *                    static int *
ncdfadir.cp     X     int *
poesadir.cp     X     time_t *                           int *
poesaget.cp     X     int *, time_t *                    int *, int *
sfcdata.c       X     time_t *                           int *
sfcmg.c         X     int *

giniutil.c      SDI   time_t * (FILELIST struct)         int *
MSAT.c          SDI   time_t *                           int *
POES.c          SDI   time_t *                           int *
SDIUtil.c       SDI   time_t *                           int *

obtgserv.cp     XCD   time_t *                           int *
wtxgserv.cp     XCD   static time_t *, time_t            static int *, int *

nowrutil.c      UPC   time_t * (FILELIST struct)         int *

amsradir.cp     XRD   int *
amsraget.cp     XRD   int *
sminadir.cp     XRD   int *
sminaget.cp     XRD   int *
swnd2cdf.c      XRD   time_t *                           int *
tminadir.cp     XRD   int *
tminaget.cp     XRD   int *

These changes are some of the modifications that need to be made to the
McIDAS CVS source code repository.

I have not yet heard back from Randy or Martha Tonkin (Martha typically does
the McIDAS installs/updates), so I don't know if everything is working for
him now.

> Thanks!

No worries.

Talk to you later...


Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
Unidata HomePage                       http://www.unidata.ucar.edu

Ticket Details
Ticket ID: ICA-762524
Department: Support McIDAS
Priority: Normal
Status: Closed

NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.